The Pump That Quits at Minute Fifteen
A 3 hp self-priming centrifugal pump lifts water 8 ft from a pond through 25 ft of 2 in. suction hose with a foot valve and a 1/4 in. mesh strainer basket. The pump feeds a sand media filter and then about 40 gpm of rotor heads at 45 psi. On a fresh start the pump primes and holds 45 psi for roughly 15 minutes; then discharge pressure falls to 22 psi, flow drops noticeably, and the pump gets loud and gravelly. After shutdown the pump will not hold prime overnight and must be reprimed every morning. The pond has been dropping all week and is now only about 6 in. above the top of the strainer.
1. Separate and explain the two failure signatures in these symptoms.
Read the worked answer
Two different failure signatures are stacked on top of each other. Separate them before touching anything.
Signature A - loses prime overnight. This is an off-cycle, static problem: water drains out of the pump case and suction line while the pump is off. Causes: a leaking foot valve (worn flapper, grit on the seat), a suction-side air leak (loose cam-lock or hose fitting, bad strainer O-ring, pinhole in the hose), or a weeping case drain plug. A suction line that drains back to an 8 ft lift refills with air every time.
Signature B - 15 good minutes, then 22 psi and a gravelly pump. This is dynamic: cavitation from a starved suction. Either the 1/4 in. mesh basket is progressively blinding with algae and pond silt, or the intake has begun to draw a vortex because the pond is only 6 in. above the strainer top. Either way, available NPSH falls, the pump cavitates, head and flow drop, and the noise follows.
The falling pond level is the likely trigger for the timing of the complaint - which also means the screen condition may be a separate, older problem.
2. Give an ordered diagnostic plan and the measurements you would take.
Read the worked answer
Work the suction side first, since both signatures point there.
- Shut down and pull the strainer. Inspect, clean, and assess open area. A 1/4 in. basket sized for 40 gpm blinds quickly in a pond; a larger basket or coarser screen is often the permanent fix.
- Look at the intake while running. Only 6 in. of submergence at 40 gpm is marginal - watch for a surface dimple indicating a vortex before the pump ever sucks air visibly.
- Vacuum gauge at the pump suction. More than about 10 in. Hg, rising as the screen blinds, confirms a restriction. A needle that fluctuates or falls toward zero points to an air leak instead.
- Check the foot valve with the pump off and the suction isolated - a leaking flapper will not hold a water column.
- Motor amps. Amps should fall as flow falls during the 15-minute decline. If amps rise, look at mechanical causes (bearing, impeller rubbing) instead of starvation.
- Filter differential. A loading sand filter also cuts flow, but it would not explain losing prime at rest. Log the filter's clean-to-dirty differential during a good run so the two causes stay distinguishable.
Record a baseline afterward: suction vacuum, discharge pressure and amps at startup.
3. The owner proposes adding a check valve at the pump discharge and dropping the strainer deeper. Evaluate both.
Read the worked answer
Discharge check valve: it helps in one narrow way - it will hold water in the pump case when the pump sits above the pond, so restarts do not require a full reprime. It does not fix a leaking foot valve (the suction line still drains and pulls air into the pump eye), and it does nothing about the cavitation. Worse, it can mask a failing foot valve until the impeller is damaged, and if any chemical injection is added later, a check valve can trap contaminated water in the line. Treat it as a crutch, not a repair.
Dropping the strainer deeper: correct in principle - more submergence kills the vortex - with conditions. Keep the intake at least 12-18 in. off the bottom so it does not dredge silt, keep the entire strainer submerged at the lowest expected pond level, and do not reduce the screen area. If the pond swings widely, a floating intake is the better answer.
The real fix is the combination: repair or replace the foot valve, find and seal the air leak, clean or enlarge the strainer, and set the intake at a depth that stays submerged. Then baseline the readings.